Rapid Visible Light-Mediated Controlled Aqueous Polymerization with

Sep 22, 2017 - Akira Watanabe , Jia Niu , David J. Lunn , Jimmy Lawrence , Abigail S. Knight , Mengwen Zhang , Craig J. Hawker. Journal of Polymer Sci...
0 downloads 0 Views 2MB Size
Letter pubs.acs.org/macroletters

Rapid Visible Light-Mediated Controlled Aqueous Polymerization with In Situ Monitoring Jia Niu,†,‡,§ Zachariah A. Page,‡ Neil D. Dolinski,‡ Athina Anastasaki,†,‡ Andy T. Hsueh,∥ H. Tom Soh,⊥,# and Craig J. Hawker*,†,‡,∥ †

California NanoSystems Institute, ‡Materials Research Laboratory, and ∥Department of Chemistry and Biochemistry, University of California, Santa Barbara, California 93106, United States § Department of Chemistry, Boston College, Chestnut Hill, Massachusetts 02467, United States ⊥ Department of Radiology, School of Medicine, and #Department of Electrical Engineering, Stanford University, Stanford, California 94305, United States S Supporting Information *

ABSTRACT: We report a simple procedure for rapid, visible lightmediated, controlled radical polymerization in aqueous solutions. Based on the photoelectron transfer reversible addition−fragmentation chain transfer (PET−RAFT) polymerization, fast chain propagation at room temperature in water was achieved in the presence of reductant and without prior deoxygenation. A systematic study correlating irradiation intensity and polymerization kinetics, enabled by in situ nuclear magnetic resonance spectroscopy, provided optimized reaction conditions. The versatility of this procedure was demonstrated through a rapid triblock copolymer synthesis, and incorporation of water-labile activated esters for direct conjugation of hydrophilic small molecules and proteins. In addition, this technique boasts excellent temporal control and provides a wide range of macromolecular materials with controlled molecular weights and narrow molecular weight distributions.

T

and in situ bioconjugation.7,29 Alternatively, the photoinitiated RAFT system developed by Cai and co-workers showed impressive reaction rates, but required deoxygenation prior to polymerization.27,30 In this report, we address this challenge through the design of simple, room temperature-based aqueous CRP procedures that reach high conversion within minutes. Optimal reaction conditions that balance propagation rate and polymerization control are systematically identified using a recently developed in situ NMR monitoring technique.31 Given that high conversion and chain-end fidelity is achievable with this platform, the rapid fabrication of triblock copolymers, and incorporation of hydrolytically unstable activated ester building blocks for in situ bioconjugation of small molecules and proteins are explored. Initial studies examined tris(2,2-bipyridyl)dichlororuthenium(II) (Ru(bpy)3Cl2) as a water-soluble photocatalyst for PET-RAFT, which, upon photoexcitation, undergoes electron/energy transfer to induce polymerization (Figure S3a).23,26 However, in this mechanism the presence of trace oxygen and the generation of singlet oxygen greatly retards chain propagation (Figure S3b).32 It was hypothesized that the addition of a reductant would mitigate this issue by preferential

he development of controlled radical polymerization (CRP) techniques has enabled the synthesis of welldefined functional polymers with predefined molecular weight and low dispersity (Đ).1,2 In particular, CRP techniques that operate in water have attracted significant attention, since they grant access to hydrophilic monomers and serve as a more environmentally friendly and biocompatible alternative to organic solvents.3−7 Nevertheless, traditional aqueous CRP techniques often rely on either metastable Cu(I) catalysts that readily undergo disproportionation in water or operation at elevated temperatures, resulting in diminished control and a loss in chain end-fidelity during the polymerization.3,4,8 Recently, there has been interest in the development of visible-light mediated CRP techniques that offer spatiotemporal control and room temperature operation.9−19 For example, Hawker, Miyake, Yagci, and others have developed visible lightmediated atom transfer radical polymerization (photoATRP) strategies,9,20−22 while Cai and Boyer expanded the platform to photoinitiated and photoelectron transfer reversible addition− fragmentation chain transfer (PET-RAFT). These advances provide polymers with low Đ and excellent chain-end fidelity.23−28 Concurrently with this work, PET-RAFT conditions were developed that allow for room temperature aqueous CRPs, though reaction rates are slow, resulting in significant irradiation times.26 Faster reaction rates are highly desired as they would reduce exposure time, making the methodology compatible with hydrolytically stable monomers © XXXX American Chemical Society

Received: August 8, 2017 Accepted: September 18, 2017

1109

DOI: 10.1021/acsmacrolett.7b00587 ACS Macro Lett. 2017, 6, 1109−1113

Letter

ACS Macro Letters

Figure 1. PET-RAFT monitored using in situ NMR spectroscopy. (a) Reaction scheme. (b) Illustration of the in situ NMR setup. (c) Reaction kinetics at different light intensities, represented by the plot of ln(M0/MT) vs reaction time (T), where M0 and MT are the concentration of monomer at time zero and T, respectively. (d) Apparent propagation rate constants (kpapp) and dispersity of the resulting polymers with respect to light intensity. (e) Temporal control during automated “ON” and “OFF” cycling of the light source. Light intensity: 26 mW/cm2.

and controlled aqueous PET-RAFT polymerizations without the need for deoxygenation, a deeper understanding of the effects of light intensity was particularly desirable for further development of this system. A digitally controlled LED coupled with in situ NMR monitoring offers the capability to simply and systematically study the effect of light intensity on these PET-RAFT systems. Tuning the light intensity (470 nm) from 8 to 139 mW/cm2 revealed a direct linear relationship with kp, while maintaining log−linear kinetics for all intensities measured as well as narrow molecular weight distributions (≲1.4) at high monomer conversion (≥85%), consistent with controlled polymerization processes (Figure 1c,d). In addition, for faster polymerization speeds at higher light intensities, induction times are shortened (Figure S8). However, it is noteworthy that a slight compromise between light intensity and Đ exists (i.e., higher intensity leads to larger Đ), as seen in Figure 1d and in the SEC traces of the resulting polymers (Figure S9), which may be attributed to an increase in side reactions at high kpapp, such as photolysis of the trithiocarbonate chain-end (Figure S10 and S11) and radical termination by chain−chain coupling. This systematic study revealed a desirable balance between kpapp and Đ at a light intensity of 26 mW/cm2, and as such, this intensity was used in all subsequent studies, unless otherwise indicated. Temporal control was then investigated using in situ NMR monitoring with automated “on”/“off” cycling of the LED (Figure 1d). Rapid chain propagation during illumination is significantly decreased when the light is turned “off” and increases quickly with the next “on” cycle with minimal inhibition and similar rates of polymerization. This “on”/“off”

reaction with singlet oxygen, while simultaneously serving as an electron source to improve catalyst turnover (Figure S3c).32−36 To this end, the polymerization of N,N′-dimethylacrylamide (DMA) as the monomer, 2-(butylthiocarbonothioyl) propionic acid (BTPA) as the chain transfer agent (CTA), and sodium ascorbate (NaAsc) as the reductant was studied under ambient conditions in water (Figure 1a). The initial experiments were screened inside an NMR spectrometer using a fiber-coupled 470 nm light emitting diode (LED; Figure 1b), revealing fast and controlled polymerization kinetics. Specifically, targeting a degree of polymerization (DP) of 400 at a light intensity of 26 mW/cm2 provided high monomer conversion (∼85%) in less than 40 min. We attributed the short induction period of ∼3 min to oxygen consumption;32,37,38 accordingly, deoxygenation of the reaction mixture prior to polymerization eliminated this initial inhibition (Figure S4). The evolution of the experimental number-average molecular weight (Mn), as measured by size exclusion chromatography (SEC), with respect to monomer conversion is linear during the course of the reaction, indicating good control over polymerization (Figure S5). Additionally, replacing NaAsc with sodium acetate (NaOAc) led to no observable polymerization over a period of 4 h (Figure S6), which attests to the importance of a reducing agent to quench singlet oxygen. The role of oxygen in polymerization inhibition was further probed via a method established by Boyer and coworkers, where anthracene-9,10-dipropionic acid disodium salt (ADPA) is used as a singlet oxygen sensitizer.32,39,40 Using this technique, we observed a significant reduction of singlet oxygen evolution in the presence of NaAsc being observed (Figure S7). While the presence of NaAsc proved critical to achieve rapid 1110

DOI: 10.1021/acsmacrolett.7b00587 ACS Macro Lett. 2017, 6, 1109−1113

Letter

ACS Macro Letters Table 1. Summary of Polymerization Conditions and Results entry

[M]/[BTPA]/ [catalyst]/ [reductant]

monomer

1 2 3 4 5 6 7

400:1:1 × 10−3:2 200:1:1 × 10−3:2 100:1:1 × 10−3:2 40:1:1 × 10−3:2 25:1:1 × 10−3:2 40:1:1 × 10−3:2 40:1:1 × 10−3:2

DMA DMA DMA DMA DMA DEA NAM

solvent water water water water water water water

+ + + +

20 20 20 20

v% v% v% v%

DMSO DMSO DMSO DMSO

catalyst concn (mM)

time (min)

αa (%)

Mn,theob

Mn,expc (SEC)

Đ

6.25 × 10−3 1.25 × 10−2 2.5 × 10−2 6.25 × 10−2 1.0 × 10−1 6.25 × 10−2 6.25 × 10−2

45 30 30 20 20 15 15

90 90 80 95 90 96 95

39900 18100 8150 4000 2460 5100 5600

43200 22100 7010 6000 1500 7020 6100

1.23 1.24 1.28 1.21 1.30 1.26 1.22

a c

Monomer conversion (α) determined by 1H NMR spectroscopy. bTheoretical molecular weight calculated based on monomer conversion. Molecular weight and Đ (Mw/Mn) determined using SEC (DMF with 0.1% LiBr as eluent, relative to PS standards). Light intensity: 26 mW/cm2.

cycling was repeated four times, exemplifying the excellent temporal control that is achievable under these conditions. To examine the versatility of this technique, the targeted molecular weights for DMA polymerizations were varied and the monomer scope was expanded. Reactions with DP ranging from 25 to 400 were conducted at a constant monomer concentration and CTA/catalyst/reductant ratio, along with light intensity (26 mW/cm2, Table 1, entries 1−5). Additionally, at low DPs (i.e., high CTA concentrations) 20 vol % DMSO was added as cosolvent to solubilize BTPA. Significantly, for a wide range of targeted molecular weights, monomer conversions of greater than 80% could be achieved within ∼15−30 min. This reaction rate is markedly higher than previously reported PET-RAFT systems, which take 5−10× longer to reach the same monomer conversion.26,37 Additionally, polymerizations of N,N′-diethylacrylamide (DEA) and Nacryoylmorpholine (NAM) with a targeted DP = 40 were successfully conducted without deoxygenation, achieving >95% monomer conversion within 15 min. Both polymers exhibited low Đ and molecular weights that were consistent with theoretical values (Table 1, entries 6 and 7). One of the hallmarks of a CRP process is the ability to generate well-defined architectures, such as multiblock copolymers. To achieve such structures, high polymer chainend fidelity is essential, which was confirmed with electrospray ionization mass spectrometry (ESI-MS) for a poly(NAM) sample prepared using the presented approach (Figure S12). To take advantage of the rapid PET-RAFT system and the resulting excellent chain-end fidelity, one-pot block copolymerizations were conducted, which mitigates the need for intermediate purification as it relies on quantitative monomer conversion for each consecutive block. First, NAM was polymerized to nearly quantitative conversion (97%) after 15 min of irradiation (26 mW/cm2), followed by DEA (97%, 15 min) and DMA (94%, 20 min). The shift to higher molecular weight after each block addition was tracked using SEC, revealing a low Đ (